Why we killed score-above-all
The single-number metric is a trap. Here's why we show you the axes and the verdict, not just the composite.
An earlier version of distro only showed the composite score. Users optimized for it. We watched the portfolio data tilt toward "safe, high-shareability, low-charge" posts — the distribution equivalent of SEO'd listicles. Everything started sounding the same.
What went wrong
The composite is a weighted sum. Like all weighted sums, it's an opinion about what matters. Once users trust it, they stop thinking about the axes. A post scoring 86 that's all shareability and no distinctiveness looks like a "winner" when it's actually forgettable.
What we changed
Three things:
- The composite is shown, but demoted visually. The axis chart is the primary display.
- The verdict is categorical (kill / refine / ship), not a rank. You decide whether to ship a 78 over an 82.
- Refinement suggestions point at specific axes. "Bump distinctiveness" not "score higher."
How to use this
Look at the axis profile, not the composite. A contrarian post might have low shareability by design; what matters is whether it has high emotional charge and distinctiveness. A data drop might have low hook strength; what matters is payoff and specificity.
If an idea has two or three strong axes and the rest are mediocre, it's a draft worth writing. If every axis is mediocre, don't bother — the composite is meaningless flat.